Pro–Con Perspectives on Ethics in Surgical Research: Update from the 39th Annual Surgical Infection Society Meeting

Opt-out

The model of consent utilized in healthcare research has substantial implications for overall research participation and the resulting ability to advance care [1,2]. Each day at every inpatient treatment facility, patients are receiving suboptimal care and are potentially being actively harmed by what care providers do not know about optimal treatment. In fact, patients are injured by what providers believe to be but is not true because the interventions and therapeutic strategies have been studied inadequately. Additionally, every day, thousands of patients willing to participate in clinical research are not provided an opportunity to do so. Unconscious bias of medical researchers alters who is approached to participate. As clinicians, we have a moral imperative to improve the care of future patients. This moral imperative provides strong rationale to pursue and utilize an opt-out, as opposed to opt-in model of consent, in appropriate settings of clinical research. Published literature supports that the opt-out consent model is superior for several reasons. The opt-out consent more closely aligns participant desires with consent, allows a greater number of potential participants to consider research, increases actual research participation, is preferred by the majority of study participants, and limits bias that the consent process introduces.

Selecting the appropriate consent models requires an assessment of balance of risk to the participant versus potential benefit of the research for the participant and future patients and the difficulty of performing the research [3]. Consent models vary related to the spectrum of this risk-benefit plus difficulty ratio, a high ratio mandates explicit consent and a very low ratio supports a waiver of consent. For areas between the two extremes, opt-out consent is most appropriate. In simple terms, the difference between opt-out and opt-in consent is as follows. After dissemination of relevant information in an opt-out model, consent is presumed, and participants actively withdraw consent. In an opt-in model, non-consent is presumed, and participants must actively provide consent. Although, the two models are presented here as absolute and discrete, gradations within each model exist (Table 1) [4].

Several lines of research support that an opt-out approach increases overall participation. Several systematic reviews of organ donation rates support that opt-out models increase organ donation rates [5,6]. Shepherd et al. [7] examined the organ donation rates over a 13-year period (2000–2012) in 48 countries with either opt-in or opt-out consent processes. Twenty-five of the 48 countries used opt-out consent processes for donation. Opt-out countries had substantially higher deceased donor rates and higher total numbers of kidneys (27% relative increase) and livers (50% relative increase) transplanted. Instrumental variables analysis suggested that the effect of opt-out versus opt-in consent was causal in the difference in the donation rates. Similarly, the opt-out consent model appears to increase participation in research. In a randomized trial of opt-out versus opt-in consent process for linkage of data to evaluate the safety of vaccines involving 1,129 parents of vaccinated children, opt-out consent resulted in 96% participation versus 21% in the opt-in consent [8]. Overall, when surveyed, 94% of parents supported data linkage and more than 90% believed that offering opt-out was either important or somewhat important.

A requirement to use opt-in consent has been shown to decrease the number of patients approached for consent and introduces bias into the research process. In a cross-sectional survey of Australian patients with cancer regarding health professional involvement in biobanking consent using an opt-in strategy, only 29% of respondents had been approached [4]. Of the 71% not approached, 59% indicated that they would have consented; an additional 12% indicated that they were unsure. Researchers also evaluated the factors associated with being approached by clinicians and found that those with post-secondary education/technical qualifications were 4.5 times more likely to be approached (odds ratio 4.5, 95% confidence interval 1.7–11.7). More than 90% of respondents indicated that opt-out consent was acceptable, whether the professional initiated the discussion or the patient initiated the discussion. Similar preferences for opt-out consent has been demonstrated for hospitalized patients. In a multicenter survey of 919 inpatients, 88% of patients either approved of or were indifferent to the collection of medical data for research and 87% approved or were indifferent to opt-out consent [9]. When asked to nominate a preferred consent method, 54% indicated a preference for opt-out consent, whereas only 28% indicated opt-in consent. Of the participants in this study, 33% indicated English as a non-preferred language. Factors found to be associated with a preference for opt-out consent included participation in prior research, English as a non-preferred language, and being illiterate in their native language. Thus, utilizing an opt-out consent process, where appropriate, increases participation, limits bias that may be introduced by healthcare professionals, and is preferred by participants in settings of limited risk.

The implications of consent models utilized are greatly magnified when medical research involves critically ill patients [10]. The severity of illness, the risk of adverse outcome, and the complexity and diversity of therapeutic interventions all complicate the performance and interpretation of research, elevating the importance of access to high-quality clinical data and adequately powered studies. The consent process in this population is particularly challenging. In a prospective, observational study of critically ill adults eligible to participate in research studies in 23 Canadian intensive care units, researchers demonstrated that 57% of all eligible patient events were missed [11]. In this population, surrogate consent was required in more than 90% of those eligible. Researchers also documented that the rationale for participation in research differed between patients and surrogate decision-makers; patients indicated that participation was driven more by the desire to benefit others and extend knowledge whereas surrogate decision-makers indicated the desire for patient benefit. Thus, one might posit that when minimal risk to the patient is present, opt-out models would align more closely with patient desires than involving surrogate decision-makers. These researchers also found clinical factors were not associated with the decision to decline consent. The likelihood of declined consent was positively associated with primary preferred language other than English and negatively associated with the experience of the research personnel involved in the consent process. Again, opt-out consent models would limit the biases introduced in this population.

For an opt-out strategy to succeed, an ethical application of principles must be applied in the planning phase. Opt-out strategies must engage and notify the potential patient pool, especially if study inclusion occurs in a short time frame. This can be achieved through deliberate and thoughtful community consultation and inclusion of community stakeholders in the planning process of the study. As health information technology advances, access to high-quality, high-volume discrete clinical data is expanding at a rapid pace. The ability to utilize this information to improve the care of all patients has great implications for future patients. Balancing the consent in favor of opt-out models is in the collective interest of these patients.

Opt-in

History is fraught with instances of medical research that violated patient autonomy and disrespected the individual right to exercise independent choice in the name of scientific discovery. In 1932, the United States Public Health Service initiated an experiment in Macon County, Alabama, to determine the natural course of untreated syphilis [12]. The men enrolled in the Tuskegee syphilis study were followed for decades, were not offered penicillin when it became widely available, and many eventually passed away from advanced syphilis. The unethical judgment of this case forms the ultimate cautionary tale in research ethics. Although we have advanced far beyond the Tuskegee syphilis trials, the concept of patient autonomy is emphasized repeatedly in modern medical research and should continue to be. Respect for patient autonomy is a critical aspect of medical ethics and research.

Informed consent requires that patients be provided with clear, detailed information, thereby allowing them to make autonomous decisions about their medical care. In the realm of research, the process of obtaining informed consent provides the opportunity for discussion and clarification about the use of patient information, giving patients the ability to control the privacy of their medical information. Opt-in versus opt-out consent processes may be guided by the estimated risk versus benefit of an intervention, as well as the difficulty of the study itself. Generally, lower risk studies without any substantial risk of harm and a strong believed potential benefit may warrant an opt-out method of consent, whereas more risky and difficult studies require explicit consent, whereby patients must actively agree to participate.

Although these ends of the consent spectrum appear clearly delineated, the reality is not as clear. Medical data are private, sensitive, and personal. The ability to manage one's own medical information and control the use of it is a valuable component of maintaining individual autonomy and promoting personal empowerment in making healthcare decisions. The use of medical data for research in an opt-out situation requires researchers to bear the responsibility for such sensitive information, without the opportunity for the patient or a surrogate to agree. Opt-out research studies require strong oversight to ensure that the study is conducted in an ethically responsible way. Conditions for overriding patient autonomy should be strict and infrequent, and violations of the patient's right to choose should occur only when necessary. Often, the concept of minimal risk is evoked when considering studies where an opt-out model of consent might be desired. Although the risk and benefit assessment is an important one, it is critical to think more broadly: the data collected should have valid scientific merit in its proposed uses, be of important public interest, and be used in an ethical, secure way. The bar needs to be set exceptionally high if patient autonomy is to be sacrificed.

Moreover, it is unclear whether opt-out strategies lead to true increases in desired outcomes. In Denmark, the national opt-out organ donation registry was implemented in 1995 and abolished in 2014; at the time of abolition, nearly 20% of the population had opted out [13]. Cited reasons for program failure include the ad hoc nature of the program without a defined purpose, the lack of a defined evaluation criteria or process, and no approach to changing attitudes over time regarding organ donation [13]. In Wales, implementation of an opt-out strategy for organ donation resulted in no substantial change in organ donation rates 15 months after the law was implemented in late 2015 [14]. A burden of proof rests on those overseeing opt-out programs to demonstrate the importance and value of the program in order to justify its paternalistic violation of patient autonomy and privacy, and even well-meaning programs such as organ donation registries may fail to prove so.

There are alternatives to opt-out research, especially for surgical research. Surrogacy consent is especially helpful in obtaining consent for those who are incapacitated, and although surrogacy decision-making is not a perfect alternative to individual autonomy, it provides an opportunity for the patient's values and wishes to be reflected. In medicine, we recognize the value and importance of individuality and the diversity of people's lives, values, and experiences. Preferentially leaning toward an opt-out system implies a paternalistic ability to understand better what is good for the patient and to predict how to use other people's private information. The potential increase in numbers of research participants is not worth the violation of patient trust and autonomy.

Pro: Patients should pay

The RCT remains the gold standard against which all other research study designs are compared. Randomization theoretically mitigates many forms of both bias and confounding that are rampant in the surgical literature. The RCT is predicated upon the principle of equipoise: the notion that there is no clear advantage to one treatment arm over the other. In fact, it is generally considered unethical to expend the resources necessary and expose patients to the risks of treatment involved in conducting an RCT for which there already exists compelling data to favor one treatment strategy over the other.

Despite these theoretical considerations, in the real world, many surgeries are undertaken routinely in the absence of the class 1 evidence provided by well-designed RCTs. Surgical stabilization of rib fractures, empiric embolization for high-grade splenic injuries in hemodynamically stable patients, and early video-assisted thoracoscopic surgery (VATS) for empyema represent common examples of this concept. In these, and many other cases, RCTs are conducted in the background of an operative intervention already occurring commonly. Furthermore, the costs associated with this operation are paid for by the patient (typically via an insurance carrier).

The issue at hand relates to the responsibility for paying for the cost of the surgical intervention in such a trial. In exploring this question, the first consideration involves a distinction between research-related and clinical costs. The first refers to ancillary tests, typically in the form of surveys, imaging, and laboratory measurement, which would only occur if the patient participates in a research trial. In the aforementioned example of early VATS for empyema, measurement of pleural fluid thromboelastography would constitute a research-related cost because its results bear no clinical consequence on the management of the patient. By contrast, whether or not the patient gets the VATS surgery is, by definition, a clinical cost, because it may have happened regardless of participation in the research trial and results in clear clinical consequences. There is no question that research-related costs should be paid for by the research study. The question at hand relates to clinical costs, and the first two reasons why these costs should be paid for by the patient are because the intervention may have been offered anyway, and the surgery has clinical consequences for the patient.

An additional issue with asking research trials to fund the costs of surgical interventions is that this reasoning fails to account for the costs of the non-operative arm of treatment. Costs of the non-operative arm of an RCT may range from additional pain medications, imaging tests, non-operative procedures (e.g., percutaneous or endoscopic procedures), and additional days in the hospital. Many of these costs are not known at the outset of the trial and may, in fact, be more expensive than the cost of the surgery itself. Asking researchers to pay for costs associated with one treatment arm but not the other is irrational.

The next concern when selectively paying for one arm of treatment is that of coercion. The U.S. Office for Human Research Protection defines coercion as the undue influence that occurs through an offer of an excessive or inappropriate reward or other overture in order to obtain compliance. In the case of surgical RCTs, informing patients that if they elect to have surgery, their operation will be paid for by someone else may unduly influence their decision to proceed with the operation. This scenario is of particular concern because most surgeries are expensive. Such enticement may disproportionately affect patients of lower socioeconomic status. Allowing for the costs of surgery to proceed by the normal workflow of third-party insurers eliminates the potential for coercion.

The issue of patient versus researcher responsibility for the payment of surgical interventions in RCTs has been adjudicated by the Center for Medicare and Medicaid Services (CMS) more than 10 years ago. According to CMS policy 310.1, the National Coverage Determination for Routine Costs in Clinical Trials, the patient is responsible for paying for all routine costs in clinical trials, including the investigational item itself, unless it is specifically covered outside of the trial. Although additional caveats apply, such as the trial having therapeutic intent, and the purpose of the trial being the evaluation of an item that falls within the Medicare benefit category, CMS has clearly sanctioned patient responsibility for these costs.

In summary, surgical RCTs compare interventions for which there is equipoise and, in most cases, involve surgeries that are already in common practice. Patients should not be held responsible for paying for research-related costs, defined as those costs that would not have occurred without participating in the RCT and for which there is no clinical benefit. Using the same logic, patients should pay for the cost of the intervention itself, as it may have occurred outside trial participation and most certainly has an effect on outcome. Furthermore, the counter argument of asking researchers to pay for surgical interventions is impractical because of costs, may constitute coercion, and is in direct opposition to established government policy.

Con: Patients should not pay

The Bayh-Dole Act of 1980 allowed industry patents to be supported by federal government-funded research, so that universities, businesses, and non-profits could hold intellectual property rights for inventions made supported by federal funding. This law attempted to increase healthcare innovation and stimulate economic growth, allowing private industries with vested interests to participate in scientific research studies and finance advances in medical care [18]. Since that time, industry has become an increasing source of research funding, with pharmaceutical, medical device, and biotechnology firms now providing the majority of medical research money [15,19,20]. With the increased role of private industry in medical research, a concern about conflicts of interests emerged.

In the early 2000s, reports started to surface regarding the potential biases associated with industry-sponsored research and bias in reporting or in publishing [21]. Early research indicated an increased likelihood of favorable publication for research funded by industry sponsors, which raised concerns regarding bias in research [21]. However, the scientific community has aggressively addressed these concerns by defining and establishing standards for ethical conduct of industry sponsored to address potential bias in research directly [22–24]. The use of Institutional Review Boards, disclosures, and ethical practices have adapted to manage potential bias in research partnerships more effectively. A 2017 meta-analysis of medical oncology research indicated that in more than 224 studies, industry-funded studies were of higher quality than non-industry–funded studies without a substantial difference in the likelihood of favorable publication.

Billions of dollars are spent annually in the United States on medical research [15,18,20]. Large sample sizes are often required to minimize type 2 errors, and RCTs are the gold standard for evidence-based medicine [25–28]. Adequate and deliberate funding allows for high-quality research design and databases, personnel to perform long-term follow-up, and well-designed RCTs. Randomized trials that are funded as “usual medical care” are at risk of lacking the resources to perform the needed design and follow-up elements adequately.

Patient-funded participation in RCTs introduces bias, even when patients are randomized, because surgical and non-surgical treatment arms do not cost the same. A population that would agree to pay for a randomized treatment arm is likely different than those who would refuse to participate based on the cost uncertainty, which limits the generalizability of the study even if participants are randomized. Pharmaceutical research does not charge patients differently for placebo medication compared with experimental treatment, because the cost differential would be unjust. Equity in clinical trials is an important consideration, because research has indicated that health outcomes vary substantially between populations [29]. Research requiring patient-funded participation introduces bias against enrollment of people of lower socioeconomic status and may result in favorable funding of studies on diseases that disproportionately affect wealthier patients.

From a philosophical perspective, research and medical care should not be conflated. Clinical research by definition does not offer a benefit of one treatment arm over the other, and patients who participate in trials must acknowledge that they are providing a public good without necessarily receiving any personal benefit [28,30]. Medical care, in contrast, is selected based on the individual benefits for a single patient. Paying for clinical research may cause confusion of the distinction between the two, introducing more bias and raising an additional question of fairness if patients cannot distinguish them.

Research is expensive, but industry, patients, and physicians all benefit from high-quality, evidence-based medicine. Effective and ethical industry-academic partnerships are both possible and desirable without placing undue financial burdens on patients.

Pro: Open-access journals are the future

The rapid evolution of the electronic media has had a major impact upon traditional publications. Many newspapers and periodicals have ceased publication because subscriptions and advertising revenues have declined rapidly because of alternative Internet sources of information. In medical and scientific publications, the loss of subscriptions, advertising, and reprint revenues have also been felt by journal publishers and it is because of institutional subscriptions that most have survived.

Other changes have impacted the traditional model of publication and subscriptions in medical and scientific journals. More and more scientific articles are published electronically, and some journals have delayed or eliminated traditional hard-copy publications. Indexing services (e.g., PubMed) have made it the practice that clinicians and scientists search and read specific articles rather than whole issues of a journal. Because the copyrights to traditional articles are owned by the publisher and not the authors, this is perceived to restrict the dissemination of new information to only those willing to pay subscription prices.

Thus, over the last decade the open-access publication model has evolved. With open access the authors submit the manuscript to their journal of choice, but if accepted for publication the author is responsible for payment of an Article Publication Charge (APC). The Gold Open-Access Agreement allows dissemination of the manuscript for non-commercial purposes. Other restrictions may apply, but publications may be available through common repositories (e.g., PubMed Central) for all to access. The Gold Open-Access Agreement is usually though a Creative Commons document with the authors that details the terms of the contract. The Green Open-Access Model is an alternative that is more restrictive and may include a six- to 24-month embargo on the release of the article for general access. During the embargo period, only journal subscribers can access the publication. A large number of variations exist in the open-access agreement, but the general result is that the publications in open-access are available for all at no cost to the reader because the cost has been borne by the authors, their funding organization, or their institution.

Most scientific journals have gone to a hybrid open-access model in which some publications within an issue are open access, and others are traditional publications that are bound by the copyright restrictions of the publisher. For hybrid journals, an increasing number of open access publications has led to some discounting of subscription prices because the authors are bearing a portion of the cost of publication. A few quality journals have gone to an exclusive open-access model (e.g., Medicine and PLOS One).

Funding agencies have now made open-access publication the wave of the future. The NIH requires that funded publications must be publicly available within three months of acceptance [31]. The International Consortium of Research Funders (cOAlition S) has offered Plan S that, among other elements, advocates that all scientific publications should be published under open-access agreements and that funders of research should not support publications in hybrid journals [32]. They recommend that monographs and book chapters be included under the open-access umbrella. They recommend sanctions by funders of those authors/institutions that are not compliant with their recommendations. Plan S is supported by the European Commission and the European Research Council and recently has been endorsed by the World Health Organization [33]. The expectation is that sponsors of research or the institution of origin (not the authors) would provide budgeted funds to support the APCs. This last expectation is highly unlikely to materialize. Although Plan S has many draconian features that will not likely be accepted, the proposal does reflect the sentiments that open access is a necessary direction for scientific publications of the future.

The future is now in medical research publication and surgeons must embrace open-access publication. Open-access publications will reduce the time between acceptance and public dissemination of research work dramatically. It will reduce the time and effort required to do literature searches for clinical information and research projects because full manuscripts will be readily available. It will greatly aid educational efforts with residents and students.

On the downside, it will be important that the academic community is vigilant in ensuring that the process is not corrupted by enhanced revenues to publishers for accepting more, and less rigorously peer-reviewed, publications. There is always the downside issue that journals with better impact factors or more prestigious reputations will expect dramatically larger APCs, especially when funding agencies or institutions are paying the tab.

Unsponsored research from efforts by surgeons and surgical residents has been a major source of important advances in surgery, and journals and academic leaders must devise methods for preserving this source of academic product when APC funding is not available. Considerations must be given for providing a waiver of APCs for research that is not sponsored but has scientific/clinical relevance. It remains unclear whether hybrid journals will persist into the future, but at present most creditable journals have endorsed a hybrid policy.

In summary, open-access publication has many positive features over the traditional methods of medical research publication. It also has vulnerabilities that must be addressed. It will be the responsibility of the academic community to ensure that the peer-review process is protected, and that open access also means open availability for all meaningful research activity to achieve publication and dissemination.

Con: Open-access journals are predatory

Open-access publishing is likely here to stay, but that does not mean that we have to be happy about it. In fact, we as investigators and authors have to protect ourselves from the numerous and rapidly growing number of journals that exist solely to generate fees as their business models without any regard to the quality and oversight of their editorial boards, or the quality and integrity of the peer-review process that is the lifeblood of medical and scientific publishing. Absent those crucial attributes, paying to publish one's own work is tantamount to purchasing an advertisement in a newspaper.

This circumstance arose after the NIH promulgated its open-access mandate in 2008, whereby all NIH-funded research had to be made available open access within 12 months. The argument proffered was that dissemination of new knowledge would be expedited and facilitated, and that investigators would receive recognition for their work—at a price. This was not alarming at the time, because the NIH had already vetted the science in general terms through its grant-review mechanism, and NIH-funded research tends to be published in high-quality refereed journals. Although new “open-access” fees were imposed by some (many?) journals, these in some cases merely substituted for the “page charges” imposed by journals that do not accept advertising. It was a wash, or nearly so.

Concurrently, such institutions often promulgated their own open access mandates, a phenomenon that has been increasing exponentially but now appears to have plateaued, as a point of saturation is reached. Whereas in 2005 only approximately 100 institutions had emplaced such a mandate, by early 2019 that number had increased more than nine-fold to approximately 950.

Several reputable publishers/journals arose or came to prominence, such as the Public Library of Science (publishers of PLoS One and other titles) and BioMed Central (owned by Springer Nature and publisher of more than 250 online titles). These respected publishers are not the problem. Their titles are listed increasingly by PubMed and Science Citation Index, making their published articles freely searchable by the scientific community at large. Searchers in research or large clinical institutions could do so at (usually) no charge because of institutional subscriptions.

The problem has become manifest because open-access journals have proliferated dramatically, with more than 11,600 listed in May 2016 by the Directory of Open Access Journals (DOAJ). Competition for publishable articles far exceeds the number of meritorious contributions, degrading quality (but no discount in fees charged), but also competition for warm bodies to contribute by “invitation” (but without waiver of fees) or to join editorial boards (as adjudged by the innumerable such solicitations flooding the e-mail in-boxes of reputable investigators). The problem is underscored by the evanescent nature of many of these journals. By December 2017, DOAJ had purged approximately 5,000 of these titles, usually for failure to respond to a query from the directory as to whether they continued to exist. Despite the purge, 12,728 journals were listed by February 25, 2019, indicating that their number had approximately doubled in three years. The scurrilous business practices of some of these journals is underscored by the incorporation of a “Black” category in a color-coding scheme to describe journals' degree of participation in open-access publishing, to denote digital piracy by widespread copyright infringement.

Currently, approximately 40% of scientific articles are published open access, ranging from a low of approximately 10% in the field of chemistry to approximately 40% in health care and 60% in biomedical research. Article “processing fees” range from a low of about $250.00 to more than $3,000.00 for Gold publications (that offer immediate, unfettered open access), with more than 450 Gold journals charging $2,000.00 or more. However, the Gold connotation does not infer anything about journal quality; currently only approximately 13% of Gold journals are listed by PubMed.

The core of the argument against open-access journals is twofold. First, there is damage to the integrity of the peer-review process and thus, to the overall quality of scientific publishing. Examples abound of deliberate hoax papers being accepted/published by open-access journals. Second, the literature is in conflict as to whether the putative benefit—more visibility/citations—is actually the case. Moreover, there may be publication bias from authors choosing to make their better science more immediately available by going the open-access route.

And what of our surgical trainees? A productive career may be stimulated to great heights by a mentor willing to polish and publish something as simple as a case report or a small case series at a formative time in the trainee's career. Publication of articles such as those may dry up altogether for lack of funds or unwillingness to pay.

Pro: Basic science research

Basic or pre-clinical science research fills a knowledge gap. It attempts to understand why things exist or how they work, often at a genetic, molecular, or cellular level. A preconceived practical application for the research is not required. The therapeutic benefits of basic science discoveries can take years or decades to materialize. Nevertheless, they provide critical insights into the mechanism of disease, and ultimately form the foundation for advances in science [34].

Clinical or applied research satisfies a predetermined need, such as whether a specific drug is safe and effective for patient use. In contrast to basic science, clinical research studies the practical application of scientific discovery. Although it can also take many years to prove the validity of basic science in the clinical arena, there is a definable end point. As such, objective-driven surgical intuition might infer that applied research is inherently more valuable. Although clinical research is essential, innumerable fundamental discoveries support basic science as the critical foundation of applied research, and therefore worthy of the majority of funding.

Technology has transformed travel. Applications such as Waze help us avoid traffic congestion in major cities. Google Maps allows us to navigate lesser known destinations confidently. Modern aircraft reach their destinations with unprecedented accuracy and precision. All of these technologies utilize global positioning systems (GPS). As GPS satellites orbit the Earth, a time difference develops between the on-board atomic clocks and those that are stationary on the Earth's surface. The time discrepancy is 38 mcsec/d, which on a GPS map equates to a distance of approximately 11 km. We can help explain and account for this difference mathematically utilizing theories (special relativity [1905] and general relativity [1915]) postulated in the early twentieth century by Albert Einstein [35]. Einstein was not interested in traffic patterns. Einstein wanted to understand how the universe worked to fill a basic knowledge gap.

With the advent of modern biologic technologies, we are starting to realize the impact of CRISPR, gene editing techniques, and the concept of precision medicine where we can manipulate the human genome to treat diseases [36,37]. We have immunotherapies that can genetically engineer T-lymphocytes to recognize and destroy tumor cells more effectively [38]. The hope is that these genetic technologies will transform how we treat cancers. Back in the 1850s, however, there was no quest to cure cancer. Rather, a basic scientist worked painstakingly to understand variety in pea plants, and why physical traits including height, seed color, and pod shape manifested in certain ways. That scientist, Gregor Mendel, is the father of modern genetics. Unlike Einstein who was credited during his lifetime, Mendel's work was ignored. It was rediscovered and replicated posthumously three decades later, illustrating the unpredictable maturation process of basic science discoveries [39].

Surgeons are well versed in the recognition and management of surgical site infections, the leading cause of health-care–acquired infections [40]. The Surgical Infection Society was founded to educate healthcare providers and the public about infections in surgical patients and promote research in the understanding, prevention, and management of surgical infections. There would be little understanding or treatment of surgical infections without the contributions of a basic scientist who observed that colonies of staphylococci on a petri dish failed to grow adjacent to mold contamination. The scientist, Alexander Fleming, then studied why bacterial growth was inhibited by the Penicillium mold. His work and that of others culminated in the world's first antibiotic substance, garnered a Nobel Prize, and saved millions of lives.

Surgeon–scientists have a long tradition of pivotal contributions to surgical basic science, and medicine in general [41]. Declining clinical reimbursement, increasing administrative and educational responsibilities, and stagnant governmental research budgets have resulted in fewer surgeons overall conducting basic research and less demonstrable success in terms of articles published and grants obtained in comparison to healthcare peers [42,43]. There is a perception that building a clinical surgery practice and pursuing basic science research are no longer compatible [44]. To reverse these trends, we must invest in the right combination of people, infrastructure, and technology to maintain basic research at the forefront of scientific discovery. In short, the majority of funding should go to basic science research.

Pro: Clinical research

Every year, the United States federal government spends more than $140 billion on research and development (R&D). Forty-eight percent of this is allocated to the Department of Defense and 44% of the remaining funds are allocated to basic science research. Approximately 23% of the funding goes to the Department of Health and Human Services (HHS), primarily the NIH (Fig. 1). This amount was just over $32 billion in 2016 [45]. Despite the large amount of funding allocated to research there are several challenges that have been difficult to overcome in the last few decades. First, a good measure for return on investment does not exist. Second, most basic science results are not translated to clinical practice. Finally, clinical research is a lengthy process partially because of lack of adequate funding and inadequate number of participants.

OPEN IN VIEWER

FIG. 1.

Federal funding for research and development in 2015. DOD = Department of Defense; HHS = Department of Health and Human Services; NASA = Department of Aeronautics and Space Administration; DOE = Department of Energy; NSF = National Science Foundation; USDA = Department of Agriculture; DOC = Department of Commerce. (Adapted from National Science Foundation statistics). Color image is available online.

At present, there is no benchmark to determine if research funding is well spent or to evaluate the effectiveness of the research being conducted. Jacob et al. [46] showed that the average NIH grant of $1.7 million is associated with an increase of only 1.2 and 2 publications in 5 and 10 years, respectively. This is an increase of only 7% compared with researchers who failed to secure NIH funding. Another criterion that might be used to determine effectiveness of research is what proportion of basic science findings are translated into a change in clinical practice. The time it takes for a new drug to be approved has increased from 7.9 years to 12.8 years since the 1960s. New drugs require more time to advance from the laboratory through clinical trials. The challenges include low patient enrollment rates and a focus on the treatment of chronic disease, which is highly cost effective [47]. Finally, life expectancy is another measure that has been used to monitor return on grant funding investment. Average life expectancy in the United States has been increasing since the 1960s, however, within the last five years it has been decreasing (Fig. 2) [48]. In summary, a single standard to measure effectiveness of research grants does not exist.

OPEN IN VIEWER

FIG. 2.

Life expectancy in the United States from 1960–2015.

Basic science research has been defined as laboratory research that does not have a practical application. The purpose of such research is to expand the understanding of a topic, not provide a comprehensive answer [49]. Over the last six decades major advances have been made in the field of basic science research. To address the challenge of implementing basic science research findings into clinical practice the new field of translational research has emerged. Translational research targets the application of basic science research results into clinical practice. In the recent years, translation has been targeted by major funding bodies. In 2005, the NIH started awarding Clinical and Translational Science Awards (CTSAs) to encourage implementation of basic research findings to patient care. However, despite billions of dollars of research grants only a small fraction of findings is turned into RCTs, and an even smaller fraction is implemented into clinical practice. One of the challenges is that basic science researchers are often dyssynchronous with the clinical setting. This results in a lack of clear methods for translation, and thus findings are lost before they reach the wider public [50].

Clinical research involves human subjects or tissue taken from humans to understand disease process. Clinical trials, epidemiologic studies, and outcomes research fall under the umbrella of clinical research. The primary challenge faced by clinical research is inadequate funding. Less than one-third of the NIH's research funding is allocated to clinical research. This figure includes animal models to study disease development and progression, thus the allocation of funding for human clinical trials is even less [49]. Another obstacle in the path of clinical research is stringent regulations imposed by governing bodies, which will ultimately need to be streamlined [51]. Furthermore, clinical research is costly, pharmaceutical companies can spend as much as $2.6 billion to market a single drug [52]. Because most clinical research is driven by pharmaceutical companies, there is a need to bridge the gap between them and basic science researchers in order to improve clinical care.

The challenges faced in conducting research today are funding, lengthy clinical trial process, and rates of translation from laboratory to clinical practice. At present, a large portion of federal grant money is awarded to the NIH, which largely funds basic science research. To implement changes into clinical practice, it is imperative that more funding is allocated to clinical research. Furthermore, there is a need for a uniform process to bring results from the bench to bedside. In addition, the gap between industry-driven research and basic science research needs to be closed. Finally, the clinical research enterprise should drive research funding and if needed clinical research collaboratives should decide on funding allocation to appropriate research.